Last week, Google quietly released an app that lets users run a range of openly available AI models from the AI dev platform Hugging Face on their phones.
Called Google AI Edge Gallery, the app is available for Android and will soon come to iOS. It allows users to find, download, and run compatible models that generate images, answer questions, write and edit code, and more. The models run offline, without needing an internet connection, tapping into supported phones’ processors.
AI models running in the cloud are often more powerful than their local counterparts, but they also have their downsides. Some users might be wary of sending personal or sensitive data to a remote data center, or want to have models available without needing to find a Wi-Fi or cellular connection.
Google AI Edge Gallery, which Google is calling an “experimental Alpha release,” can be downloaded from GitHub by following these instructions. The home screen shows shortcuts to AI tasks and capabilities like “Ask Image” and “AI Chat.” Tapping on a capability pulls up a list of models suited for the task, such as Google’s Gemma 3n.
Google AI Edge Gallery also provides a “Prompt Lab” users can use to kick off “single-turn” tasks powered by models, like summarizing and rewriting text. The Prompt Lab comes with several task templates and configurable settings to fine-tune the models’ behaviors.
Your mileage may vary in terms of performance, Google warns. Modern devices with more powerful hardware will predictably run models faster, but the model size also matters. Larger models will take more time to complete a task — say, answering a question about an image — than smaller models.
Google’s inviting members of the developer community to give feedback on the Google AI Edge Gallery experience. The app is under an Apache 2.0 license, meaning it can be used in most contexts — commercial or otherwise — without restriction.
Keep reading the article on Tech Crunch
Meta recently released a new video editing app for creators called Edits. The new app is designed to rival ByteDance’s CapCut, a popular short-form video-editing app used by many creators.
Meta first shared that it was working on Edits back in January after ByteDance-owned CapCut was removed from U.S. app stores when the TikTok ban momentarily went into effect. Although the app has since come back online and is available to download, TikTok’s future in the U.S. remains uncertain, so Meta is ready to capitalize on CapCut’s possible absence in the future with its new Edits app.
We’ve created this guide to walk you through the app and the different features it offers, and we provide an overview of how it compares to CapCut.
Edits is available to download on both iOS and Android. Once you download the app and open it, you will be asked to log in with your Instagram account.
From there you will see five main tabs: Ideas, Inspiration, Projects, Record, and Insights. The Ideas tab offers creators a space to jot down ideas for new videos and save reels to a collection, and the Inspiration tab features a hub for discovering trending audios and other popular reels.
The Projects tab is where creators can store and revisit the videos they currently have in progress. This tab is also where creators can upload videos from their phone’s camera roll. The Record tab lets creators start filming video content right within Edits. Lastly, the Insights tab gives creators an overview of how their content is performing by surfacing metrics like views, reach, and follower counts.
Green screen: Replace and edit the background of your videos with one tap.
Timeline: Precisely arrange and adjust clips.
Captions: Add automatic captions to your videos in multiple languages.
Audio library: Add music from Instagram’s library to your videos.
Cutouts: Isolate specific objects with precision tracking.
Animate: Turn static images into videos using AI.
Apply all: Apply filters, effects, transitions, and adjustments to all of your clips at once.
Timeline frame rate selector: Convert your video to the frame rate you want in order to enhance playback smoothness.
Alignment guides: Ensure viewers can see the important parts of your video when posted and accurately position text, emoji, or other elements.
Beat markers: Add auto-detected beat markers to help you align clips, text, and overlays with audio when editing.
Filters: Choose from 30 different filter options to add a unique look to your videos.
Transitions: Choose from 30 different transitions to connect different video clips in a seamless way.
While both Edits and CapCut are great options for editing short-form videos, there are some differences between the two services.
CapCut offers more robust editing options, including AI features, than Edits does. Of course, this can be expected since Edits is a lot newer than CapCut. Meta will likely continue to build out Edits over the next months and years, so we can expect to see more advanced functionality in the future. It’s also worth noting that CapCut has a more extensive music library compared to Edits.
Another difference is the fact that Edits currently doesn’t have a subscription offering (for now), but CapCut does. While CapCut offers a free version, some of its advanced tools are locked behind a paid subscription. Edits is free to use, but this will likely change in the future, as Instagram head Adam Mosseri has stated that later versions of Edits may include paid features.
In addition, while CapCut is available on the web, Edits is mobile-only. However, this could change in the future.
Keep reading the article on Tech Crunch
WordPress.com parent company Automattic is changing direction… again.
In a blog post titled “Returning to Core” published Thursday evening, Automattic announced it will unpause its contributions to the WordPress project. This is despite having said only last month that the 6.8 WordPress release would be the final major release for all of 2025.
“After pausing our contributions to regroup, rethink, and plan strategically, we’re ready to press play again and return fully to the WordPress project,” the new blog post states. “Expect to find our contributions across all of the greatest hits — WordPress Core, Gutenberg, Playground, Openverse, and WordPress.org. This return is a moment of excitement for us as it’s about continuing the mission we’ve always believed in: democratizing publishing for everyone, everywhere,” it reads.
Automattic says it’s learned a lot from the pause in terms of the many ways WordPress is used, and that it’s now committed to helping it “grow and thrive.”
The post also notes that WordPress today powers 43% of the web.
It’s unclear what has changed.
However, according to sources who spoke to TechCrunch, Automattic CEO Matt Mullenweg explained in an internal post published last night that he wants to get a 6.9 release out this year with an admin refresh and something from the “new AI team.” (The latter references this week’s announcement that WordPress formed a new team to steward the development of AI projects.)
“I don’t think that will happen without our contributions,” the post stated, according to a source familiar with the internal discussions.
Insiders are speculating there’s more to it than that, and various theories are being floated. Some wonder if Mullenweg was pressured into this move or if he realized that pulling back wasn’t good for his business or its reputation? Others are debating if this change of heart has to do with Automattic’s ongoing litigation with WordPress’ hosting company, WP Engine, which Automattic has called a “threat” to the WordPress community and a “cancer to WordPress.”
No one knows what to believe, and none of these backchannel theories line up with the officially stated reason.
Automattic was asked for additional comment.
Since 2024, Automattic has been engaged in a legal dispute that has to do with how little, in Mullenweg’s opinion, WP Engine contributes to the WordPress project, despite its size and revenue.
He sees the hosting company as profiting off the open source work WordPress is doing without giving back. Mullenweg also alleges that WP Engine benefits from the confusion between WordPress and commercial services like WP Engine. This led him to ban the company from accessing WordPress.org and sue in court for unauthorized trademark usage.
Simply put, Mullenweg thinks WP Engine should either pay a direct licensing fee or up its contributions to the open source WordPress project, or shouldn’t be allowed to use its trademark.
WP Engine responded that it doesn’t think it needs a license and that Automattic misunderstands trademark law, suing Automattic in return.
Last month, Automattic laid off 16% of staff, saying the restructuring was necessary to be more agile and improve its productivity and profitability.
Sarah Perez can be reached at @sarahperez.01 on Signal and sarahp@techcrunch.com .
Keep reading the article on Tech Crunch
X’s new DM feature, XChat, has begun rolling out to beta testers. Some X users who pay for the platform’s subscription product have reported that they can access XChat, while reverse engineer Nima Owji confirmed to TechCrunch that the new messaging system appears to be ready to ship.
XChat is intended to be a more robust version of X’s existing DM (direct messages) feature, which is a holdover from before Elon Musk’s acquisition of Twitter.
Some features that have been rumored to be part of XChat include group messages, end-to-end encryption, vanishing mode, the ability to mark messages as unread, and file sharing. Messages also appear to be secured behind a four-digit passcode.
X announced on Thursday that it was pausing work on its encrypted DM feature, which could be related to the impending release of XChat, since this new system would render the existing DM interface obsolete. Paid subscribers on X have had access to a limited version of encrypted messaging for two years.
In the years since he bought Twitter, Musk has spoken on numerous occasions about his ambitions to make a Signal-like messaging service embedded into X. If XChat really is gearing up for public use, then this news could represent Musk’s follow-through on that goal.
Keep reading the article on Tech Crunch
We’re thrilled by the overwhelming response to our call for speakers at TechCrunch Disrupt 2025, taking place October 27–29 at Moscone West in San Francisco.
After a careful selection process, we’ve narrowed it down to 20 impressive finalists—10 breakout sessions and 10 roundtables.
Now it’s your turn to help shape the agenda. Audience Choice voting is open until 11:59 p.m. PT tonight. This is your final opportunity to weigh in—vote for your favorite sessions and help decide which will take the stage. You can vote for as many sessions as you’d like—just one vote per session.
The top 5 in each category will join the official Disrupt 2025 lineup.
Breakout Sessions
How to Get Acquired in Tech (Without Selling Out): M&A Tips for Founders and Builders
Aklil Ibssa, Head of Corporate Development and M&A, Coinbase
Agentic AI for Startups: Automate, Adapt, and Accelerate Growth
Anmol Rastogi, Head of Product, Amazon Business – AI & ML, Amazon
Automation with Agents: From Work Enablement to Work Completion
Chet Kapoor, Chairman and CEO, DataStax
AI at the Brink: Strategic Playbook for National Security
Dan Hendrycks, Executive and Research Director, Center for AI Safety (CAIS)
Leading a Series A Round in 2025 and Sustaining Momentum
Gabriel Kra, Managing Director, Prelude Ventures
The Agentic Apocalypse: Securing the Enterprise in the Age of 1 Billion AI Agents
Jack Hidary, CEO, SandboxAQ
Jim Breyer, Founder and CEO, Breyer Capital
Embracing AI for a Better Digital Future
Matt Madrigal, Chief Technology Officer, Pinterest
Mining for Millions with GenAI’s 4 Ds: Striking Trust, Delight, and Dividends
Michael Stewart, Managing Partner, M12
From Data to Agents: Building the AI-Native Enterprise
Sridhar Ramaswamy, Chief Executive Officer, Snowflake
From Vibes to Velocity: How AI Tools Can Help You Achieve Your Development Goals
Tim Rogers, Staff Product Manager, GitHub Copilot, GitHub
Roundtable Sessions
Future of Space Economy in the Low Earth Orbit
Abhijeet Kumar, Invited Lecturer – New Space Economy, UC Berkeley | Tech and Strategy Lead, Archer
From Startup to Scale-Up: A GTM Blueprint
Anjai Lal, Head of Strategy and Enablement, Google Cloud
From Code to Capital: How VCs Spot the Next AI Powerhouses
Avi Bharadwaj, Investment Director, Intel Capital
The Winning Formula: Turning Your Business Into a Trusted, Scalable Community To Drive Growth
Justine Palefsky and Tasneem Amina, Co-founders, Kindred
Vlad Loktev, Partner, Index Ventures
How to Train Your Model: Taming AI Agents Without Breaking Them
Kyla Guru, Head of Model Cyber Policy, Anthropic
Alex Moix, Investigations Lead, Safeguards, Anthropic
Going a Layer Deeper: Why the future of AI investments lies with infrastructure and applications
Paul Drews, Managing Partner, Salesforce Ventures
Scaling Search and AI for Millions: Lessons from Reddit Search
Rachel Miller, Product Manager, Reddit
AI Evaluation 101: Addressing Challenges to Real-World AI Applications
Rohit Patel, Director, Generative AI, Meta
From Workarounds to Breakthroughs: How UpLink Lets Users Connect Any App—No Integration Needed
Scott Weinert, CTO and Co-founder, Atomic
Whose Company Is It, Anyway?: What You Lose When You Accept Outside Capital
Sridhar Vembu, Chief Scientist, Zoho
Keep reading the article on Tech Crunch
June 5 is almost here — bringing real, unfiltered AI conversations… and higher ticket prices. Lock in your savings now.
Register now to save $300 on your TechCrunch Sessions: AI pass — and get 50% off for your +1. Don’t wait for rates to spike when event doors open.
Join us at UC Berkeley’s Zellerbach Hall — the one-day epicenter for next-gen AI insights, big questions, and actionable ideas from the builders, thinkers, and investors shaping the future.
Explore the full agenda here.
Main stage and breakout sessions are packed with tactical insight and bold vision from leaders like:
Whether you’re pitching your AI startup, swapping ideas with fellow builders, or just getting started, networking at TC Sessions: AI is smarter — thanks to the Braindate app. Use the app to match on topics, meet face-to-face, and make meaningful connections with people who care about what you care about.
And when the event’s a wrap? The conversations keep flowing at Side Events all week long in Berkeley.
This isn’t another hype-filled AI event. It’s where the noise drops out — and the real conversations begin.
Only 6 days left to lock in your low ticket rate for TechCrunch Sessions: AI. Don’t sit this one out — save $300 on your pass, and get 50% off for your +1.
Register now before prices jump at the door, and be part of the conversations actually shaping the future of AI.
Interested in a deeper discount? Participate in our AI trivia for a chance to purchase a ticket at $200 and receive a second ticket for free.
Keep reading the article on Tech Crunch